Multi-institutional development and external validation of machine learning-based models to predict relapse risk of pancreatic ductal adenocarcinoma after radical resection

Journal of Translational Medicine

Litao Yang¹  Junjie Huang²  Yiqun Fan³  Jianyao Lou⁴  Xiawei Li⁴  Yulian Wu⁴  Aiguang Shi⁴  Mingchen Zhao⁵  Zheping Yuan⁵ 

[1] Department of Surgery, Cancer Hospital of the University of Chinese Academy of Sciences (Zhejiang Cancer Hospital), 310000, Hangzhou, Zhejiang, China;Department of Surgery, Changxing People’s Hospital, 313100, Huzhou, Zhejiang, China;Department of Surgery, Fourth Affiliated Hospital, Zhejiang University School of Medicine, 322000, Yiwu, Zhejiang, China;Department of Surgery, Second Affiliated Hospital, Zhejiang University School of Medicine, 310000, Hangzhou, Zhejiang, China;Key Laboratory of Cancer Prevention and Intervention, China National Ministry of Education, Cancer Institute, Second Affiliated Hospital, Zhejiang University School of Medicine, 310000, Hangzhou, Zhejiang, China;Cancer Center, Zhejiang University, 310058, Hangzhou, Zhejiang, China;Hessian Health Technology Co., Ltd, 100007, Beijing, China;

关键词: Machine learning;    PDAC;    Relapse;    Prediction model;    Radical surgery;   

DOI  :  10.1186/s12967-021-02955-7

来源: Springer

【 摘 要 】

BackgroundSurgical resection is the only potentially curative treatment for pancreatic ductal adenocarcinoma (PDAC) and the survival of patients after radical resection is closely related to relapse. We aimed to develop models to predict the risk of relapse using machine learning methods based on multiple clinical parameters.MethodsData were collected and analysed of 262 PDAC patients who underwent radical resection at 3 institutions between 2013 and 2017, with 183 from one institution as a training set, 79 from the other 2 institution as a validation set. We developed and compared several predictive models to predict 1- and 2-year relapse risk using machine learning approaches.ResultsMachine learning techniques were superior to conventional regression-based analyses in predicting risk of relapse of PDAC after radical resection. Among them, the random forest (RF) outperformed other methods in the training set. The highest accuracy and area under the receiver operating characteristic curve (AUROC) for predicting 1-year relapse risk with RF were 78.4% and 0.834, respectively, and for 2-year relapse risk were 95.1% and 0.998. However, the support vector machine (SVM) model showed better performance than the others for predicting 1-year relapse risk in the validation set. And the k neighbor algorithm (KNN) model achieved the highest accuracy and AUROC for predicting 2-year relapse risk.ConclusionsBy machine learning, this study has developed and validated comprehensive models integrating clinicopathological characteristics to predict the relapse risk of PDAC after radical resection which will guide the development of personalized surveillance programs after surgery.

【 授权许可】

CC BY